US2008279298A1PendingUtilityA1

Multiple-input multiple-output (mimo) detector incorporating efficient signal point search

40
Assignee: COMSYS COMM & SIGNAL PROC LTDPriority: May 10, 2007Filed: May 10, 2007Published: Nov 13, 2008
Est. expiryMay 10, 2027(~0.8 yrs left)· nominal 20-yr term from priority
H04L 25/03318H04L 25/03242H04L 2025/03414H04L 2025/03605
40
PatentIndex Score
0
Cited by
0
References
0
Claims

Abstract

A novel and useful apparatus for and method of multiple input multiple output (MIMO) detection for use in MIMO based communication systems. The mechanism of the invention performs a simplified tree search utilizing a single stage expansion of the most likely first symbol candidates, in the case of a 2×2 MIMO system. The invention also provides a refinement mechanism that is operative to significantly improve the log likelihood (LLR) of the list of candidates. To improve the LLR, the mechanism applies refinements rounds to generate additional candidates for both first and second detected symbols.

Claims

exact text as granted — not AI-modified
1 . A method of symbol detection for use in multiple-input, multiple-output (MIMO) systems, said method comprising the steps of:
 choosing a plurality of K candidates for a first symbol;   choosing a single second symbol to be mated with each of said K first symbol candidates to form K contender symbol pairs;   wherein each second symbol in each contender symbol pair is chosen to yield an optimum cost function when mated with its respective first symbol; and   wherein K is a positive integer.   
   
   
       2 . The method according to  claim 1 , wherein each of said K first symbol candidates is chosen from a first symbol constellation. 
   
   
       3 . The method according to  claim 1 , wherein K equals 16. 
   
   
       4 . The method according to  claim 1 , wherein K equals 25. 
   
   
       5 . The method according to  claim 1 , wherein said constellation comprises 64 Quadrature Amplitude Modulation (QAM) symbols. 
   
   
       6 . The method according to  claim 1 , wherein said plurality of K first symbol candidates are chosen as the set of K distinct values of s 1  which substantially minimizes the following expression
   ∥y 1 −R 11 s 1 ∥ 2      
     wherein s 1  denotes the first symbol candidate, y 1  denotes a received first symbol after QR decomposition and R 11  denotes a channel estimate after QR decomposition. 
   
   
       7 . The method according to  claim 1 , wherein said cost function for choosing a best symbol mate for said each first symbol candidate comprises 
     
       
         
           
             
               
                 s 
                 2 
               
                
               
                 ( 
                 i 
                 ) 
               
             
             = 
             
               
                 
                   arg 
                    
                   
                       
                   
                    
                   min 
                 
                 
                   s 
                   2 
                 
               
                
               
                 
                    
                   
                     
                       y 
                       2 
                     
                     - 
                     
                       
                         R 
                         21 
                       
                        
                       
                         
                           s 
                           1 
                         
                          
                         
                           ( 
                           i 
                           ) 
                         
                       
                     
                     - 
                     
                       
                         R 
                         22 
                       
                        
                       
                         s 
                         2 
                       
                     
                   
                    
                 
                 2 
               
             
           
         
       
     
     wherein s 1  denotes said first symbol candidate, s 2 (i) (i=1 . . . K) denotes said i th  second symbol candidate, y 1  and y 2  denote received first and second symbols after QR decomposition, respectively, and R 21  and R 22  denote channel estimates after QR decomposition. 
   
   
       8 . The method according to  claim 1 , wherein said plurality of K first symbol candidates are found directly using binary search techniques. 
   
   
       9 . The method according to  claim 1 , wherein a second symbol candidate is found for each of said first symbol candidates using binary search techniques. 
   
   
       10 . The method according to  claim 1 , wherein said plurality of K first symbol candidates are found directly using slicing techniques. 
   
   
       11 . The method according to  claim 1 , wherein a second symbol candidate is found for each of said first symbol candidates using slicing techniques. 
   
   
       12 . A detection apparatus for use in multiple-input, multiple-output (MIMO) systems, comprising:
 a processor operative to:
 choose a plurality of K candidates for a first symbol; 
 choose a single second symbol to be mated with each of said K first symbol candidates to form K contender symbol pairs, wherein each second symbol in each contender symbol pair is chosen to yield an optimum cost function when mated with its respective first symbol; and 
   a memory for storing said K candidate symbol pairs.   
   
   
       13 . The apparatus according to  claim 12 , wherein each of said K first symbol candidates is chosen from a first symbol constellation. 
   
   
       14 . The apparatus according to  claim 12 , wherein K equals 16. 
   
   
       15 . The apparatus according to  claim 12 , wherein K equals 25. 
   
   
       16 . The apparatus according to  claim 12 , wherein said constellation comprises 64 Quadrature Amplitude Modulation (QAM) symbols. 
   
   
       17 . The apparatus according to  claim 12 , wherein said plurality of K first symbol candidates are chosen as the set of K distinct values of s 1  which substantially minimizes the following expression
   ∥y 1 −R 11 s 1 ∥ 2      
     wherein s 1  denotes the first symbol candidate, y 1  denotes a received first symbol after QR decomposition and R 11  denotes a channel estimate after QR decomposition. 
   
   
       18 . The apparatus according to  claim 12 , wherein said cost function for choosing a best symbol mate for said each first symbol candidate comprises 
     
       
         
           
             
               
                 s 
                 2 
               
                
               
                 ( 
                 i 
                 ) 
               
             
             = 
             
               
                 
                   arg 
                    
                   
                       
                   
                    
                   min 
                 
                 
                   s 
                   2 
                 
               
                
               
                 
                    
                   
                     
                       y 
                       2 
                     
                     - 
                     
                       
                         R 
                         21 
                       
                        
                       
                         
                           s 
                           1 
                         
                          
                         
                           ( 
                           i 
                           ) 
                         
                       
                     
                     - 
                     
                       
                         R 
                         22 
                       
                        
                       
                         s 
                         2 
                       
                     
                   
                    
                 
                 2 
               
             
           
         
       
     
     wherein s 1  denotes said first symbol candidate, s 2 (i) (i=1 . . . K) denotes said i th  second symbol candidate, y 1  and y 2  denote received first and second symbols after QR decomposition, respectively, and R 21  and R 22  denote channel estimates after QR decomposition. 
   
   
       19 . The apparatus according to  claim 12 , wherein said plurality of K first symbol candidates are found directly using binary search techniques. 
   
   
       20 . The apparatus according to  claim 12 , wherein a second symbol candidate is found for each of said first symbol candidates using binary search techniques. 
   
   
       21 . The apparatus according to  claim 12 , wherein said plurality of K first symbol candidates are found directly using slicing techniques. 
   
   
       22 . The apparatus according to  claim 12 , wherein a second symbol candidate is found for each of said first symbol candidates using slicing techniques. 
   
   
       23 . An apparatus for multiple-input, multiple-output (MIMO) detection in a digital receiver, comprising:
 means for determining a plurality of best candidate first symbols from among a constellation of symbols; and   means for matching each candidate first symbol with a second symbol to produce a list of contender symbol pairs, wherein each second symbol in each contender symbol pair is chosen to yield an optimum cost function when mated with its respective first symbol.   
   
   
       24 . The apparatus according to  claim 23 , wherein said optimum cost function comprises a partial distance function. 
   
   
       25 . The apparatus according to  claim 23 , wherein said optimum cost function comprises a partial distance function in accordance with the following 
     
       
         
           
             
               
                 s 
                 2 
               
                
               
                 ( 
                 i 
                 ) 
               
             
             = 
             
               
                 
                   arg 
                    
                   
                       
                   
                    
                   min 
                 
                 
                   s 
                   2 
                 
               
                
               
                 
                    
                   
                     
                       y 
                       2 
                     
                     - 
                     
                       
                         R 
                         21 
                       
                        
                       
                         
                           s 
                           1 
                         
                          
                         
                           ( 
                           i 
                           ) 
                         
                       
                     
                     - 
                     
                       
                         R 
                         22 
                       
                        
                       
                         s 
                         2 
                       
                     
                   
                    
                 
                 2 
               
             
           
         
       
     
     wherein s 1  denotes said first symbol candidate, s 2 (i) (i=1 . . . K) denotes said i th  second symbol candidate, y 1  denotes a received first symbol, y 2  denotes a received second symbol and R 21  and R 22  denote channel estimates. 
   
   
       26 . The apparatus according to  claim 23 , further comprising means for selecting a best symbol pair from said the contender symbol pairs on said list that yields a minimum Euclidean distance as a hard decision. 
   
   
       27 . A computer program product characterized by that upon loading it into computer memory a multiple-input, multiple-output (MIMO) detection process is executed, said computer program product comprising:
 a computer usable medium having computer usable program code for performing detection in a MIMO based digital receiver, said computer program product including;   computer usable program code for choosing a plurality of K candidates for a first symbol, wherein K is a positive integer; and   computer usable program code for choosing a single second symbol for each of said K first symbol candidates, wherein said second symbol is chosen to yield a symbol pair comprising a first symbol and second symbol that optimize a cost function.   
   
   
       28 . The computer program product according to  claim 27 , further comprising computer usable program code for selecting a symbol pair from among said plurality of K candidate symbol pairs as a hard decision. 
   
   
       29 . A multiple-input, multiple-output (MIMO) radio receiver coupled to a plurality of antennas, comprising:
 a radio frequency (RF) receiver front end circuit for receiving a plurality of radio signals transmitted over a MIMO channel and downconverting the received radio signals to baseband signals;   a demodulator adapted to demodulate said baseband signal in accordance with the modulation scheme used to generate said transmitted radio signals;   a MIMO detector processor operative to:
 choose a plurality of K candidates for a first symbol; 
 choose a single second symbol for each of said K first symbol candidates, wherein said second symbol is chosen to yield a symbol pair comprising a first symbol and a second symbol that optimizes a cost function; 
 generate bit log likelihood ratio (LLR) values as a function of said K candidate symbol pairs; and 
   a channel decoder operative to receive and decode said bit LLR values to generate receive data therefrom.   
   
   
       30 . The MIMO radio receiver according to  claim 29 , wherein said MIMO detector processor is further operative to select a symbol pair from among said plurality of K candidate symbol pairs as a hard decision. 
   
   
       31 . A method of multiple-input, multiple-output (MIMO) detection in a digital receiver, said method comprising the steps of:
 determining a plurality of best candidate first symbols from among a constellation of symbols;   matching each candidate first symbol to one or more values of a second symbol to produce a list of contender symbol pairs, wherein each second symbol in each contender symbol pair is chosen to yield an optimum cost function when mated with its respective first symbol; and   selecting a best symbol pair from the contender symbol pairs on said list.   
   
   
       32 . A method of multiple-input, multiple-output (MIMO) detection in a digital receiver, said method comprising the steps of:
 determining a plurality of candidate first symbols from among a first constellation of symbols in a first layer;   at each successive layer k, spanning every candidate symbol in a k th  constellation of symbols to find the best M k  candidate symbols to constitute a next layer k+1;   pairing a single symbol match in a last layer N T  to each candidate of a second to last layer N T −1; and   wherein k denotes the layer, M k  denotes the number of candidate symbols spanned, N T  denotes the dimension of a transmitted symbol vector.   
   
   
       33 . A method of multiple-input, multiple-output (MIMO) detection, said method comprising the steps of:
 determining an optimal candidate symbol pair, comprising the steps of:
 choosing a plurality of K candidates for a first symbol; 
 choosing a single second symbol to be mated with each of said K first symbol candidates to form K contender symbol pairs; 
 wherein each second symbol in each contender symbol pair is chosen to yield an optimum cost function when mated with its respective first symbol, wherein K is a positive integer; 
 choosing a best symbol pair comprising a first detected symbol and second detected symbol from among said K contender symbol pairs; 
   refinement of said second detected symbol, comprising the steps of:
 for each information bit of said second detected symbol, searching for a constellation symbol having a bit of interest opposite that of a corresponding bit of said second detected symbol that yields a minimum Euclidean distance to said second detected symbol; and 
 forming additional candidate symbol pairs, each comprising a first detected symbol mated with one of said found constellation symbols.

Cited by (0)

No later patents cite this yet.

References (0)

No backward citations on record.